Phosphorus mineralization kinetics and response of microbial phosphorus to drying and rewetting in a Florida Spodosol

Surface soils of a north central Florida Spodosol (sandy, siliceous hyperthermic Alaquod) from fertilized and unfertilized plantations of loblolly pine (Pinus taeda L.) were conditioned, dried, rewet and incubated at 38°C for up to 26 d with periodic sampling for inorganic P and microbial P. Undried samples were also incubated and sampled periodically. Several kinetic models were evaluated to describe patterns of net P mineralization. Cumulative net mineralization of P in undried samples of both fertilized and unfertilized treatments was best described by zero-order kinetics. In contrast, a segmented model with two pools was most appropriate for describing cumulative net mineralization of P in dried and rewet samples, with one pool following zero-order kinetics, and the other following first-order kinetics. Net mineralization in rewet soils progressed in three stages: (i) An initial flush where inorganic P was brought into solution, the source most likely being turnover of the microbial biomass from the previous drying period and mineralization of organic substrates; (ii) a lag of a few days where there was no net release of P; and (iii) a period that followed similar kinetics to the undried soil, where the microbial biomass had recovered sufficiently to mineralize P from soil organic matter. Although more P was mineralized in fertilized soils, the kinetics of the reactions were similar to those in unfertilized soils. Generalized models were used to predict net P mineralization expressed as a percent of total P (specific P mineralization) for all data (both fertilized and unfertilized treatments). A zero-order model best described mineralization of P in undried soils (R2=0.884), and a segmented two-pool model was the best fit for soils that had been dried and rewet (R2=0.923). There was a negative relationship between inorganic P (as a % of total P) and microbial P in undried soils (R2=0.715). When soils were dried and rewet, microbial P increased over the entire incubation period but this relationship fluctuated with time and was not significantly correlated with P mineralization. The kinetic models proposed here should be useful in improving predictions of P mineralization in other sandy soils of low adsorption capacity.